Course: Studying Healthy Psychosislike Experiences to Improve Illness Prediction
CME Credits: 1.00
Released: 2023-03-08
Abstract
Importance Distinguishing delusions and hallucinations from unusual beliefs and experiences has proven challenging.Observations The advent of neural network and generative modeling approaches to big data offers a challenge and an opportunity; healthy individuals with unusual beliefs and experiences who are not ill may raise false alarms and serve as adversarial examples to such networks.
Conclusions and Relevance Explicitly training predictive models with adversarial examples should provide clearer focus on the features most relevant to casehood, which will empower clinical research and ultimately diagnosis and treatment.
To identify the key insights or developments described in this article
View Full Course